Genomic prediction using preselected DNA variants from a GWAS with whole-genome sequence data in Holstein–Friesian cattle

Abstract

BackgroundWhole-genome sequence data is expected to capture genetic variation more completely than common genotyping panels. Our objective was to compare the proportion of variance explained and the accuracy of genomic prediction by using imputed sequence data or preselected SNPs from a genome-wide association study (GWAS) with imputed whole-genome sequence data.MethodsPhenotypes were available for 5503 Holstein–Friesian bulls. Genotypes were imputed up to whole-genome sequence (13,789,029 segregating DNA variants) by using run 4 of the 1000 bull genomes project. The program GCTA was used to perform GWAS for protein yield (PY), somatic cell score (SCS) and interval from first to last insemination (IFL). From the GWAS, subsets of variants were selected and genomic relationship matrices (GRM) were used to estimate the variance explained in 2087 validation animals and to evaluate the genomic prediction ability. Finally, two GRM were fitted together in several models to evaluate the effect of selected variants that were in competition with all the other variants.ResultsThe GRM based on full sequence data explained only marginally more genetic variation than that based on common SNP panels: for PY, SCS and IFL, genomic heritability improved from 0.81 to 0.83, 0.83 to 0.87 and 0.69 to 0.72, respectively. Sequence data also helped to identify more variants linked to quantitative trait loci and resulted in clearer GWAS peaks across the genome. The proportion of total variance explained by the selected variants combined in a GRM was considerably smaller than that explained by all variants (less than 0.31 for all traits). When selected variants were used, accuracy of genomic predictions decreased and bias increased.ConclusionsAlthough 35 to 42 variants were detected that together explained 13 to 19% of the total variance (18 to 23% of the genetic variance) when fitted alone, there was no advantage in using dense sequence information for genomic prediction in the Holstein data used in our study. Detection and selection of variants within a single breed are difficult due to long-range linkage disequilibrium. Stringent selection of variants resulted in more biased genomic predictions, although this might be due to the training population being the same dataset from which the selected variants were identified.Electronic supplementary materialThe online version of this article (doi:10.1186/s12711-016-0274-1) contains supplementary material, which is available to authorized users.

Multi-omics-data-assisted genomic feature preselection improves the accuracy of genomic prediction

“…We found that using S_GWAS did not improve the prediction accuracy values, especially for p-value cutoffs less than 0.001 (Table 2). Similar results were also indicated in previous studies using SNPs preselected from GWAS (Veerkamp et al, 2016;Ye et al, 2019). The main reason for this result is that overfitting decreases the prediction accuracy.…”

“…Therefore, pre-selected potential causal markers or QTLs from WGS are necessary for improving the accuracy of genomic prediction (Raymond et al, 2018). Thus, many preselection variant strategies were used to improve the power of genomic prediction based on the following methods: genomewide association study (GWAS) (Zhang et al, 2014;Veerkamp et al, 2016;Song et al, 2019;Ye et al, 2019), Bayesian procedures (Kemper et al, 2015), genome-wide signatures of selection (Ye et al, 2020), Animal QTLdb (Song et al, 2019), gene annotation (Heidaritabar et al, 2016;Gao et al, 2017), and gene ontology categories (Edwards et al, 2016;Abdollahiarpanahi et al, 2017). These methods mainly depend on the direct link between phenotype and DNA variants or some prior genome annotation information.…”

Hybrid peeling for fast and accurate calling, phasing, and imputation with sequence data of any coverage in pedigrees

“…The accuracy of 35 genomic selection and the power of genome wide association studies depend on both the number 36 of individuals that have genomic data and its density (e.g., Daetwyler et al, 2008;Hayes et al, 37 2009; Hickey et al, 2014; Gorjanc et al, 2015) . The goal is then to generate genomic data on as 38 many individuals as possible at as high of a density as possible with the upper limit being the 39 presence of whole genome sequence on hundreds of thousands or millions of individuals (Hickey,40 2013; Daetwyler et al, 2014;Veerkamp et al, 2016). 41…”

Prospects of whole-genome sequence data in animal and plant breeding

“…However, when the number of imputed variants was reduced, and those were selected based on association with the trait, then the peaks eventually were found (Calus et al, 2016). Similar, when single SNP testing was applied (Veerkamp et al, 2016) in the same data, clearer QTL peaks were identified when using sequence data in comparison with SNP array data. Other studies showed that a multi-breed population results in more precise QTL mapping than a single breed population (Raven et al, 2014;Kemper et al, 2015).…”

“…Pre-selection can be performed by first performing a GWAS and select the SNPs with the lowest p-values for genomic prediction. Veerkamp et al (2016) investigated this approach, but did not find a clear benefit of using whole-genome sequence data for genomic prediction using the same data as used in Chapter 4.…”